Two strains of rat have been developed by selective breeding: one strain (R rats) is resistant to salt hypertension, the other strain (S rats) is highly susceptible. The inheritance of these traits has been explored in the first (F₁) and second (F₂) generation of crossbred rats and in backcrosses between parent and first filial (F₁ x R, F₁ x S) generations. Male F₁ rats had an average blood pressure close to the mid-parental (R and S) values, and the average of F₂ males was equivalent to that of F₁. Male offspring of F₁ with R, or F₁ with S also showed averages close to the respective mid-parental values. Female offspring showed deviations from this linear relationship, indicating a significant dominance in the female for the genes of normal blood pressure. A model of two autosomal, nonlinked diallelic loci, with a dominance deviation at one locus in the female, gave predictions with a reasonable agreement to the observed values. The same model also appeared compatible with human data if we assume a gene frequency of 0.13 for the hypertensinogenic allele on both loci. Random fluctuations in blood pressure, and incomplete homogeneity of parental strains permit several alternative models. The major conclusions are: that more than one locus is needed to explain the findings though as few as two loci may possibly suffice; the allelic effect seems additive in males, but there is a sex-determined influence on the expression in females; there is no consistent evidence for sex-linked inheritance. Furthermore, this model developed from the study of rats may provide a framework for analysis of human data.